Intelligent Control by Unfalsified Control Theory

In this paper, we show how far modern control techniques could achieve intelligent control. It depends on how much knowledge a computer can learn plant's behavior from ob-served data. Intelligent control system could mean an intelligent control algorithm must detect true performances of all candidate controllers especially when they are not in the feedback loop. In other words, there is no chance for a bad candidate to be selected to perform in the loop. In this work, we analyze the true stabilizing region and then we demonstrate one way to estimate it via unfalsified control theory. Finally, we compare them and both optimized controllers are computed using a linear case study.